Opera 6 & 7 scrolling.

**Randy Webb** · Jul 20 '05, 02:07 PM

Re: Opera 6 & 7 scrolling.

Pete wrote:
[color=blue]
> Therefor, if there isn't a better scroll method?, I need an Opera 6/7
> detection. I know that object detection is prefered over browser
> detection but I don't of any differences to reference between 6 and 7.[/color]

Look at the userAgent string. I am sure that Opera 6 doesn't contain
Opera 7.23 in it (which is what my Opera 7 has in it). No matter which
spoof mode I had it in, it contained that snippet.

As well, I am sure there has to be some feature difference between them,
other than the userAgent string.

--
Randy
Chance Favors The Prepared Mind
comp.lang.javas cript FAQ - http://jibbering.com/faq/

**Richard Cornford** · Jul 20 '05, 02:07 PM

Re: Opera 6 & 7 scrolling.

"Pete" <someonesgotmyn ame2002@yahoo.c o.uk> wrote in message
news:1abcadcc.0 401271947.13442 bf3@posting.goo gle.com...
<snip>[color=blue]
>Anyway, if there's enough page content to cause scrolling and
>I want the objects (colored divs) to scroll with the mouse in
>Opera 6. I give the mouse y coords (-) window.pageYOff set and
>the objects (+) window.pageYOff set.
>
>However, in Opera 7 it doesn't work. Only the following objects
>need (+) window.pageYOff set. The mouse code doesn't need anything
>and it works as I want?
>
>Therefor, if there isn't a better scroll method?, I need an
>Opera 6/7 detection. I know that object detection is prefered
>over browser detection but I don't of any differences to
>reference between 6 and 7.[/color]

The simple positioning of CSS position:absolu te; DIV elements taking
scrolling into account does not need significantly different handling
between Opera 6 and 7 (apart, possibly from the handling of appending
CSS units to the top and left values).

It is much morel likely that your problem is with the differences in the
reporting of mouse co-ordinates between the two. With Opera 6 only
providing clientX and clientY values but reporting those values as if
they were pageX and pageY, While Opera 7 returns genuine client area
co-ordinates for those values and also provides pageX and pageY as
alternatives.

When reading mouse co-ordinates I prefer the pageX and pageY values and
only fall back to clientX and clientY if they are missing, such as on
IE, IceBrowser and Opera 6. But knowing that Opera 6 values do not need
adjusting by the scroll values I use the presence of the window.opera
object as indicating that the clientX/Y values do not need adjusting by
the scroll (and [root].clientTop/Left) values in order to be equivalent
to the page values. That is the nearest that I have come to needing to
do anything resembling browser detecting.

Richard.

**Pete** · Jul 20 '05, 02:07 PM

Re: Opera 6 & 7 scrolling.

Randy Webb <hikksnotathome @aol.com> wrote in message news:<uZSdnYO7x 8Xrq4rd4p2dnA@c omcast.com>...
[color=blue]
> Look at the userAgent string. I am sure that Opera 6 doesn't contain
> Opera 7.23 in it (which is what my Opera 7 has in it). No matter which
> spoof mode I had it in, it contained that snippet.
>
> As well, I am sure there has to be some feature difference between them,
> other than the userAgent string.[/color]

Yes, I tried that and like you got the required result with all Opera
7.22 spoofs - see code below. However, I don't have versions 5 or 6 to
test with and will it be OK with future versions? Oh well, thanks
anyway. Unless some one comes up with something better I'll use it for
now. Like I said, it's only a bit fun.
Pete.

<script type="text/javascript">
//Main thing------------------------
var ua=navigator.us erAgent;
var op=(ua.indexOf( "Opera") != -1);
var op7;
var oa=0;
var ob=0;
var oc=0;
var ov=0;
if (op){
oa=ua.indexOf(" Opera")+6;
ob=oa+4;
oc=ua.substring (oa,ob);
ov=parseFloat(o c);
}
op7=(op && ov >= 7)?true:false;
//--------------------------------

//Test----------------------------
var a=(op)?"Yes":"N o";
var b=(op)?"Is it version 7 or higher\? ":"";
var c;
if (op && ov >= 7) c="Yes";
else if (op && ov < 7) c="No";
else c="";
var d=(op)?"Its\' version " +ov:"";
alert('Is it Opera\? '+a+'\n'+b+c+'\ n'+d);
//--------------------------------
</script>

**Richard Cornford** · Jul 20 '05, 02:07 PM

Re: Opera 6 & 7 scrolling.

"Pete" <someonesgotmyn ame2002@yahoo.c o.uk> wrote in message
news:1abcadcc.0 401280710.70f4c e58@posting.goo gle.com...
<snip>[color=blue]
>Yes, I tried that and like you got the required result with
>all Opera 7.22 spoofs - see code below. However, I don't
>have versions 5 or 6 to test with and will it be OK with
>future versions? Oh well, thanks anyway. Unless some
>one comes up with something better I'll use it for now.
>Like I said, it's only a bit fun.[/color]
<snip>

Being able to tell the difference between two versions of Opera
regardless of their spoofing modes does not solve the problem as you
still cannot tell the difference between Opera and another browser
spoofing Opera. There was a time when I don't think any other browser
would spoof Opera (couldn't see much point) but I was wrong and there
are browsers out there that offer an option to spoof the Opera UA string
in their preferences.

But the script you are describing can be (and has been) implemented in a
cross-browser form without the need to be interested in the UA string at
all. You would probably benefit more in the long run by finding out what
it is you are doing wrong that is provoking you apparent problem and
fixing that than setting off in a direction (UA string testing) that has
already been demonstrated to be worthless.

Richard.

**Pete** · Jul 20 '05, 02:08 PM

Re: Opera 6 & 7 scrolling.

"Richard Cornford" <Richard@litote s.demon.co.uk> wrote in message news:<bv8b4c$6t k$1$8300dec7@ne ws.demon.co.uk> ...[color=blue]
> The simple positioning of CSS position:absolu te; DIV elements taking
> scrolling into account does not need significantly different handling
> between Opera 6 and 7 (apart, possibly from the handling of appending
> CSS units to the top and left values).[/color]

I wouldn't have thought so either but it's what I'm getting. Maybe it
would be better if I include the script. It's stripped down just for
Opera and the pageYOffset thing.

Copy and paste some more <p>'s to cause scrolling then alternate the
pageYOffset's. As is - it works in 7 but not 6 and vice-versa.

<html><head><ti tle>test</title>
<style type="text/css">
#box{
position:absolu te;top:0px;left :0px;
width:20px;heig ht:20px;backgro und:blue;
}
</style>
</head><body>

<div id="box"></div>

<script type="text/javascript">
var y1=x1=y2=x2=y3= x3=0;
var temp=document.g etElementById(" box").style;

function mouse(){
y1 = event.clientY;// - window.pageYOff set;
x1 = event.clientX;
}
document.onmous emove=mouse;

function move_these(){
temp.top = y2 + window.pageYOff set;
temp.left = x2;
}

function delay(){
y2 = y3 += (y1 - y3) * 0.1;
x2 = x3 += (x1 - x3) * 0.1;
move_these();
setTimeout('del ay()',50);
}

window.onload=d elay;
</script>

<p> </p>
<p> </p> more....

</body>
</html>

[color=blue]
> It is much morel likely that your problem is with the differences in the
> reporting of mouse co-ordinates between the two. With Opera 6 only
> providing clientX and clientY values but reporting those values as if
> they were pageX and pageY, While Opera 7 returns genuine client area
> co-ordinates for those values and also provides pageX and pageY as
> alternatives.[/color]

I didn't know that. Would..
((event.clientY )||(event.pageY &&event.clientY ))
?event.clientY: event.clientY-window.pageYOff set;
...do?
[color=blue]
> When reading mouse co-ordinates I prefer the pageX and pageY values and
> only fall back to clientX and clientY if they are missing, such as on
> IE, IceBrowser and Opera 6. But knowing that Opera 6 values do not need
> adjusting by the scroll values I use the presence of the window.opera
> object as indicating that the clientX/Y values do not need adjusting by
> the scroll (and [root].clientTop/Left) values in order to be equivalent
> to the page values. That is the nearest that I have come to needing to
> do anything resembling browser detecting.
>
> Richard.[/color]

**Richard Cornford** · Jul 20 '05, 02:08 PM

Re: Opera 6 & 7 scrolling.

"Pete" <someonesgotmyn ame2002@yahoo.c o.uk> wrote in message
news:1abcadcc.0 401281748.30dce 4f7@posting.goo gle.com...
<snip>[color=blue]
>I wouldn't have thought so either but it's what I'm
>getting. Maybe it would be better if I include the
>script.[/color]

It almost always is.

<snip>[color=blue]
>function mouse(){
> y1 = event.clientY;// - window.pageYOff set;
> x1 = event.clientX;
>}[/color]

There you go, Opera 6 returns co-ordinates relative to the page for
clientX/Y while Opera 7 (correctly) returns co-ordinates relative to the
client area.
[color=blue]
> document.onmous emove=mouse;[/color]

Netscape 4 (but preferably not later Netscape versions (though they do
support the function)) requires calls to document.captur eEvents if you
want to follow mouse movement at the document level.
[color=blue]
> function move_these(){
> temp.top = y2 + window.pageYOff set;
> temp.left = x2;
> }[/color]

If the mouse co-ordinates you are using are page relative you do not
need to adjust for the scrolling values at this point, if you use client
co-ordinates then you do need to adjust.

But the CSS units value comes in at this point. On standards compliant
browsers a string value should be assigned to the top and left style
properties with appended units. OTOH older browsers prefer the values to
be provided as numbers. My strategy for dealing with this problem is
based on using typeof to test one of style.top or style.left to see if
it is of string type or numeric. If it is a string then it is best to
provide CSS units, if it is a number then it is best to assign a number.
I do this by assigning a value to a suitably scoped variable based on
the typeof test:-

var posMod = (typeof temp.top == 'string')?'px': 0;

-and then when a value is to be assigned to top or left (or any other
style value that need similar handling) I use the + operator:-

temp.top = y2 + posMod;

- the - y2 - value is a number and - posMod - is either the number zero
or it is the string 'px' so the action of the + operator depends on the
type of posMod. If - posMod - is zero then - + - does addition and
adding zero to a number results in an unchanged number, and that is
assigned to the style property. But if - posMod - is the string 'px'
the - + - operator does string concatenation, type-converting the - y2 -
value to a string and then appending 'px' to it, providing the desired
string value with the CSS units.
[color=blue]
> function delay(){
> y2 = y3 += (y1 - y3) * 0.1;
> x2 = x3 += (x1 - x3) * 0.1;
> move_these();
> setTimeout('del ay()',50);
> }[/color]

If - delay - always calls - delay - (without the option of being
cancelled) it would be wroth looking into the setInterval function as an
alternative.

<snip>[color=blue]
> I didn't know that. Would..
> ((event.clientY )||(event.pageY &&event.clientY ))
> ?event.clientY: event.clientY-window.pageYOff set;
> ..do?[/color]

Over the years I have tried out various strategies for reading the mouse
co-ordinates, the following is the latest version of my cross browser
document.onmous emove following object:-

var MoveFollow = (function(){
var global = this;
var theOne = {
getPageX:functi on(){
return ((readScroll.sc rollLeft|0) -
(readScroll.cli entLeft|0) + XY.x);
},
getPageY:functi on(){
return ((readScroll.sc rollTop|0) -
(readScroll.cli entTop|0) + XY.y);
},
getPageXY:funct ion(){
return {x:this.getPage X(),y:this.getP ageY()};
},
upDate:function (ev){
mm(ev);
}
};
var XY = {x:0,y:0};
var readScroll = {scrollTop:0,sc rollLeft:0,
clientLeft:0,cl ientTop:0};
var eventInited = false;
var posReadX = 'pageX';
var posReadY = 'pageY';
function mm(e){
e = e||global.event ;
XY.x = e[posReadX];
XY.y = e[posReadY];
};
function initEvent(ev){
ev = ev||global.even t;
if(ev){
if(typeof ev.pageX != 'number'){ //Opera 7 has pageX
posReadX = 'clientX';
posReadY = 'clientY';
if(!global.oper a){ //Not an Opera browser
if((typeof document.compat Mode == 'string')&&
(document.compa tMode.indexOf(' CSS') != -1)&&
(document.docum entElement)){
readScroll = document.docume ntElement;
}else if(document.bod y){
readScroll = document.body;
}
}
}
setUpOnMouse(mm );
mm(ev);
eventInited = true;
}else{
setUpOnMouse(in itEvent);
}
};
function setUpOnMouse(f) {
if(document.onm ousemove != f){
document.onmous emove = f;
if((document.ca ptureEvents)&&( global.Event)&&
(document.layer s)&&(typeof Layer == 'function')){
document.captur eEvents(Event.M OUSEMOVE);
}
}
};
initEvent();
return (function(){
return theOne;
});
})(); //simultaneously define and call (one-off)!

The - MoveFollow - global variable is assigned a function and once it
has been assigned any other code can call the - MoveFollow - function,
which will return a reference to an object that represents a common
interface to the onmousemove events (cross-browser). The bulk of the
testing and configuration work is done as this script initialises itself
so once it is running it is very efficient.

The reference to the interface is acquired as:-

var mouseMoveFollow er = MoveFollow();

-the the latest mouse X and Y co-ordinates (relative to the HTML page)
are retrieved using the interface's getPageX/Y methods:-

var pageXcord = mouseMoveFollow er.getPageX();
/* and */
var pageYcord = mouseMoveFollow er.getPageY();

If you can understand the code (with its inline function expression
call, nested functions, and so on) you should be able to understand how
the code decides which values to read from the event objects. It is the
only code that I use that has any interest in which browser is on use
specifically in order to cope with the incorrect reporting of clientX/Y
in Opera 6. Which is done by checking Opera's global - opera - property.
That isn't a problem for Opera 7, which has the pageX/Y properties on
the event object and the code prefers to use those if available.

Richard.

**Pete** · Jul 20 '05, 02:10 PM

Re: Opera 6 & 7 scrolling.

Richard Cornford:

Thanks for the script and the obvious effort put in to it, you
certainly know your onions. Also, thanks for introducing me to the
"typeof". Had loads of fun with that and very handy. I see what you
mean about pageY vs. clientY in Opera 7. pageY/X definitely from now
on. This eliminated my original 6/7 problem.

I do know about the document.captur eEvents etc for Netscape 4 but
deliberately didn't write for it. It would be nice to write for older
browsers but I think you have to draw the line somewhere. Nothing of
importance depends on my efforts so I'm just going to concentrate on
the present, future and, hopefully, standards.

Anyway, although yours is excellent, I never feel right if I haven't
put the effort in and written something myself so here's my 'lesser'
effort. I managed to fix up my old PC with 6 and the script works OK
and on the new one with Explorer 6, Netscape 7.1 and Opera 7.22 with
various and no doctypes.

One thing I didn't understand was your setInterval over setTimeout but
I'll look that up my self.
Thanks again,
Pete.

Doctype just for show.

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html
PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
<head>
<title>Modern </title>
<style type="text/css">
#box{
position:absolu te;top:0px;left :0px;
width:20px;heig ht:20px;backgro und:blue;
}
</style>
</head>
<body>

<div id="box"></div>

<script type="text/javascript">

</script>

<p> </p><p> </p>
Lots more....

</body>
</html>

Opera 6 & 7 scrolling.

Opera 6 & 7 scrolling.

Comment

Comment

Comment

Comment

Comment

Comment

Comment